This application addresses the mandates of the RFA by proposing to develop a highly novel and previously untested technology to identify, assess, and compare immunogens for cancer vaccines, employing an in vivo immune system in a high throughput screening mode. The development of the new technology will bring together multiple investigators experienced in a wide variety of complementary techniques, from different institutions and disciplines. If successful, the new technology will have great impact on oncology by providing investigators with an important new technology to evaluate and compare immunogens for cancer vaccines. We will develop and then employ this new technology in a test case to identify immunogens that elicit immune responses to the survivin tumor-associated antigen, an attractive and widely studied tumor vaccine candidate useful for this developmental proof-of-principle study due to its demonstrated activity in cancer vaccine models and relatively small size. The technology exploits an intact GI mucosal immune system as a massively parallel in vivo screening device. To develop this new screening technology we will: 1) Create a library of bacteria transformed with DNA bar-coded plasmids that include expression cassettes that place large amounts of chimeric protein, which will include sequences encoding an exhaustive, overlapping library of peptides derived from survivin, on the Gram-negative bacterial surface, 2) Feed the library to mice, 3) Use PhyloChip microarray technologies (or high throughput sequencing) to identify, via the barcodes, members of the library showing decreased relative abundance in the mouse feces over time, which we would take as evidence of the induction of a mucosal immune response against the chimeric proteins encoded by plasmids showing decreased abundance. We will rescreen the clones to confirm their ability to induce an immune response and evaluate sera and cells from animals inoculated with bacteria expressing the proteins and animals immunized with the proteins identified in the screen to determine if they induce anti-survivin humoral and cellular immune responses. Clones and the peptides encoded by the clones able to induce an anti-survivin response will be studied in an in vivo tumor vaccine model and compared to peptides already known to induce an anti-tumor immune response in the model. The study proposes to develop an innovative, rapid, high-throughput approach to the identification of potential immunogens useful as cancer vaccine employing the survivin tumor associated antigen as a model. If successfully developed, the technology also would be useful in the search for other immunogens for cancer vaccines and for comparing the immunogenicity of different candidate proteins/peptides. Since the proteins are expressed in Gram-negative bacteria, including vaccine strains of Salmonella, the technology could be used to rapidly and cheaply produce a candidate vaccine. Although the project is risky, the project could therefore potentially have a very large impact on cancer immunotherapy.